Right, so remember on Friday when, after the latest CFTC data hit, we noted that specs were the most net long TY since 2007?
And remember what we sarcastically said about that? Here’s a reminder:
…if you’re the type of person who has been conditioned to believe that “smart” money positioning may be a contrarian indicator, then it might be time to short Treasurys (again).
Yeah, well even if you are that “type of person,” you might not be the “type of person” who trusts financial commentary written by a pseudonym based on a TV show character, so we thought maybe we’d let the smartest guy on the Street tell you the same thing.
Here’s the latest from said smartest guy on the Street, Deutsche Bank’s Aleksandar Kocic…
Via Deutsche Bank
Current spec positioning data is showing substantial overweight at the long end. This is a result of short covering that started in mid January and had essentially two legs (six weeks each). The most recent leg corresponds to the last two weeks. In terms of its size, it is a multiple standard deviation — a 1.6-sigma long in the 17-year sample and post-2008, 2.8-sigma. In terms of monthly adjustments, this looks even more extreme: It is the largest change in terms of short covering (about 3.5-sigma), contested in size only with the switch to underweight immediately after the 2016 elections. the two figures show the magnitude and 1M change in TY positioning.
In contrast, positioning in the belly has hardly changed in the last month. The market is still slightly underweight there, and close to neutral at the front end.
Typically, excessive overweight implies subsequent sell off and vice versa. Spec positioning data, both levels and changes, is a contrarian indicator of subsequent rate changes. The forecasts based on positioning are usually noisy, but give relatively reliable signals regarding direction of rates, typically with hit ratios around 70%. The reliability of the signal improves with the size of over or underweight – the extremes have higher hit ratios.
When taken in its least committal form, these data are suggestive of little resistance towards bear steepening of the 5s/10s sector, consistent with our macro view of higher risk premia in the curve and in tune with other measures of risk premia across different markets.